Georgios Exarchakis is a Research Scientist (Chargé de recherche) at the Institut Hospitalo-Universitaire (IHU) Strasbourg, specializing in Machine Learning and Theoretical Neuroscience. Previously, he held postdoctoral positions at Institut de la Vision and Prof. Stephane Mallat’s DATA team at École Normale Supérieure. He earned a Dr. rer. nat. in Machine Learning from the University of Oldenburg (2016), an M.Sc. in Computational Science from Goethe University Frankfurt (2012), and a Mathematics Diploma from Aristotle University of Thessaloniki (2008). His research focuses on sparse coding, graphical models, signal processing, and deep learning, with applications in spike sorting, quantum chemistry, and natural image statistics. He developed the Kymatio library for wavelet scattering transforms and contributed to ProSper, a probabilistic sparse coding library. His work bridges theoretical neuroscience and machine learning, emphasizing interpretable models and scalable algorithms. Key contributions include advancing variational sampling techniques for generative models, solid harmonic wavelet scattering for molecular property prediction, and discrete sparse coding frameworks. He teaches machine learning courses at École Polytechnique and Oldenburg University, emphasizing practical applications and theoretical foundations. Georgios collaborates with interdisciplinary teams, including the Redwood Center for Theoretical Neuroscience at UC Berkeley and the University of Frankfurt, to advance methodologies in computational neuroscience and AI. His current research at IHU Strasbourg integrates machine learning into image-guided surgery technologies.
